How to write a nursing care plan

The nursing process in five stages

ADPIE is the framework taught across the UK, Australia, New Zealand, Canada and the US, with local variations in terminology. It is a cycle rather than a list: evaluation feeds back into assessment, and a care plan that never changes is a care plan nobody is using.

Each stage answers a question. What is going on? What is the patient's response to it? What are we aiming for? What will we do? Did it work?

In an assessed care plan, the marks are distributed across all five, and the two that students under-weight are diagnosis and evaluation — the first because it is conceptually unfamiliar, the second because it comes last and gets rushed.

Assessment

Systematic, not impressionistic. Use the framework your programme teaches — Roper-Logan-Tierney's activities of living in much of the UK, Gordon's functional health patterns elsewhere, or an ABCDE approach in acute settings.

Collect subjective data — what the patient reports, in their words — and objective data — observations, measurements, test results. Distinguish them clearly, because they carry different weight and a care plan that presents a patient's report as an observation is misleading.

Include the patient's own priorities. A plan built entirely on clinical observations without the patient's stated concerns will be marked down in any programme teaching person-centred care, and rightly.

Cite your assessment tools. A validated pain scale, a falls risk assessment, a nutritional screening tool — each has a source and a validation study, and naming it shows the assessment was structured rather than improvised.

Diagnosis — the stage that trips people up

A nursing diagnosis is not a medical diagnosis. The doctor diagnoses pneumonia; the nurse diagnoses the patient's response to it — impaired gas exchange, activity intolerance, anxiety, risk of infection.

The standard structure is three parts, known as PES: the problem, the aetiology, and the signs and symptoms. 'Impaired gas exchange related to alveolar-capillary membrane changes as evidenced by SpO2 of 89% on room air, respiratory rate 28, and reported breathlessness on minimal exertion.'

The 'related to' clause must be something nursing can act on. 'Related to pneumonia' is not useful because nursing cannot treat pneumonia directly; 'related to alveolar-capillary membrane changes' points at interventions that are within scope.

A risk diagnosis has only two parts, because the problem has not occurred: 'risk of falls related to unsteady gait and sedating medication'. There is no 'as evidenced by', since evidence of the problem would mean it had already happened.

Writing goals that can be evaluated

Goals are the patient's, not the nurse's, and they must be measurable, because the evaluation stage depends on it. 'Patient will maintain SpO2 above 94% on 2L oxygen by end of shift' can be evaluated; 'patient will breathe better' cannot.

Use the SMART structure, and note that the T matters most in a care plan — short-term goals are typically shift or day-based, long-term goals span the admission.

Write the goal in terms of what the patient will do or achieve, not what the nurse will do. 'Patient will ambulate 20 metres with a frame twice daily' is a goal; 'nurse will assist with mobility' is an intervention.

Every diagnosis needs at least one goal, and every goal must be traceable back to a diagnosis. A plan with three diagnoses and one goal has two unaddressed problems.

Interventions and rationales

Each intervention is a specific nursing action with a frequency, and each is paired with a rationale explaining why it works, supported by a citation. The rationale column is where the academic marks are.

Where the evidence comes from

Rationales need citations, and the hierarchy matters. National clinical guidelines outrank textbooks; textbooks outrank websites; and your local trust or health service policy sits alongside national guidance and sometimes overrides it.

Use guidelines from the relevant jurisdiction. NICE guidance in the UK, the Australian Commission on Safety and Quality in Health Care, the New Zealand Ministry of Health, and their equivalents elsewhere — the recommendations differ and a marker in your country will notice a plan built on foreign guidance without comment.

Cite systematic reviews and Cochrane reviews for effectiveness questions. A single trial is weaker evidence for an intervention than a review of trials, and knowing that distinction is part of what is assessed.

Keep the evidence current. Nursing guidance is revised frequently, and a rationale citing a superseded guideline is a substantive error rather than a formatting one.

Implementation

In an academic care plan, implementation is usually described rather than performed, and the description should be specific: who does what, when, and what is documented.

Note delegation explicitly where relevant. What can be delegated to a healthcare assistant, what requires a registered nurse, and what requires escalation are all decisions with a professional basis, and stating them shows you understand scope of practice.

Document communication. Handover, escalation to medical staff, and referral to other professions are nursing interventions and belong in the plan.

Evaluation — the stage that gets rushed

Evaluation asks whether the goal was met, and it must be answered against the goal as written. If the goal was SpO2 above 94% by end of shift, the evaluation states the SpO2 at end of shift.

Three outcomes: met, partially met, not met. Each requires a next step. A goal that was met may be discontinued or advanced; one that was not met requires the plan to be revised, and the revision is part of the evaluation.

Say why. 'Goal not met; SpO2 remained at 90-91% despite positioning and oxygen therapy, suggesting the underlying consolidation has not resolved. Escalated to medical team; plan revised to include chest physiotherapy referral.' That paragraph demonstrates clinical reasoning, which is what the assessment is for.

A care plan that ends with 'goal met' for every goal and no revisions reads as a plan written backwards from a tidy conclusion.

Confidentiality

Never use a real patient's name, date of birth, hospital number, or any detail that could identify them, and say in your introduction that details have been anonymised or that the case is composite.

Your regulator's code applies to academic work. The NMC in the UK, AHPRA and the NMBA in Australia, and the Nursing Council in New Zealand all treat breaches in submitted coursework as fitness-to-practise matters, not academic ones.

Anonymise the setting too. Naming the ward and the trust alongside a clinical detail can identify a patient even when the name is removed.

Common failures

Six, and all are visible before the content is read.

• A medical diagnosis in the nursing diagnosis position.
• A 'related to' clause that nursing cannot act on.
• Goals with no measurable criterion.
• Interventions with no rationale, or rationales with no citation.
• Evaluation that does not measure against the goal as written.
• Any identifying patient detail.